Commentary: The Continued Role and Value of Imaging for Acute Ischemic Stroke

The authors1 provide a comprehensive overview of the various imaging modalities that have played and continue to play an integral role in the triage of patients with acute ischemic stroke. As they mention, the cornerstone of imaging evaluation of these patients remains noncontrast computed tomography (CT) of the head, which serves to confirm eligibility for intravenous tissue plasminogen activator, excludes many stroke mimics, and helps determine eligibility of patients with large-vessel occlusion (LVO) for endovascular treatment and guide prognosis through the development of scores, such as the Alberta Stroke Programme Early CT Score (ASPECTS).2 As the authors of this manuscript discuss, advances in imaging techniques provide further information and data for all of these components. The use of perfusion imaging modalities has become widely accepted as necessary prior to endovascular intervention for stroke. Although the idea of ASPECTS was revolutionary, radiographic assessments have evolved with newer trials, including DWI or CTP Assessment with Clinical Mismatch in the Triage of Wake-Up and Late Presenting Strokes Undergoing Neurointervention with Trevo (DAWN),3 suggesting that a better understanding of collateral flow may be warranted. Although the results of the Interventional Management of Stroke (IMS) III trial suggest that the quality of collaterals predicts functional outcome,4 an understanding of how other clinical and radiographic variables affect long-term outcome after stroke thrombectomy is still needed. In addition, recent articles have focused on the importance of groin access to recanalization times as well as groin puncture time from time of stroke symptom onset.5 Although CT perfusion (CTP) imaging, magnetic resonance (MR) imaging, and the evaluation of collaterals through CT angiography (CTA) provide a more sophisticated assessment of a given patient's prognosis and response to stroke intervention, a growing body of literature is focusing on the benefits of a more efficient preintervention imaging protocol.6 While noncontrast head CT is a screening tool for the identification of many stroke mimics, CTA and CTP may be equally important to help guide management and informed discussions about intervention with patients and their families. Patients undergoing advanced neuroimaging modalities (eg, perfusion imaging) have been shown to have improved long-term outcomes, suggesting that the information from recent mechanical thrombectomy studies does impact clinical decision-making.5,7 With the advent of RAPID imaging (iSchemaView Inc, Menlo Park, California) with automated CTP, the focus has been on effective triage, diagnosis, and assessment of appropriate patients into angiographic suites.8,9 However, automated CTP, similar to CTP evaluated by physicians, has the potential to overestimate core volume, and a correlation with noncontrast imaging is becoming increasingly recognized.8,10 This inaccuracy is particularly true in the hyperacute settings after stroke symptom onset. Similar imaging paradigms are being developed for MR imaging, although studies are demonstrating that screening duration and image capture time inherently add extra time to the imaging model.5,6 Advances even in understanding and automating the evaluation of noncontrast head CT offers promise.11 Relative noncontrast CT maps can show areas of early ischemic change, which correlate with territories that functionally improve after thrombectomy.11 Although CT map studies have been conducted in small cohorts, protocols to incorporate vessel imaging as well as functional mapping of noncontrast head CT may be an efficient protocol to improve stroke symptom onset to reperfusion times. With increasingly sophisticated imaging capabilities, how we use them in clinical practice continues to evolve. The current use of advanced imaging modalities, such as CTP and MR perfusion, and the concept of deficit–infarct mismatch are tools of exclusion. In other words, physicians use these imaging concepts to identify patients who should not proceed for endovascular intervention because they are unlikely to gain benefit or may be at increased risk ofhemorrhagic conversion. Furthermore, as mentioned previously, the imaging findings are also used to guide discussions with patients and families; for example, a patient who appears to have a 50% core infarct volume will likely not do as well as a patient with a 5% core infarct volume despite recanalization. As the authors of this article allude, this begs the question – should we delay groin stick (ie, endovascular intervention) to obtain perfusion imaging if the imaging is a source of exclusion? Expanding this concept brings the physician to an ethical crossroad. If we elect not to proceed with endovascular intervention because perfusion imaging suggests there is a large core infarct and the patient is unlikely to gain benefit from recanalization, we accept the patient's poor outcome as a function of the natural course of his or her disease process. Conversely, if we forgo perfusion imaging in this same patient, proceed with endovascular intervention, and obtain complete recanalization (thrombolysis in cerebral infarction grade 3), but the patient experiences a large reperfusion hemorrhage, we then accept the patient's poor outcome at our hands. Even the ASPECTS may have little influence on outcome following mechanical thrombectomy. Bhatt et al. found no statistically significant difference in outcomes between patients with ASPECTS 0 to 5 and those with ASPECTS 6 to 7 undergoing mechanical thrombectomy for LVO.12 Therefore, one can imagine a situation in which patients who have symptoms suggestive of stroke and negative head CT (performed simply to exclude hemorrhage or stroke mimic) proceed directly to the angiography suite, with a diagnosis of vessel occlusion made through digital subtraction angiography (DSA) rather than noninvasive means (eg, CTA). Then, if LVO is seen, recanalization is pursued. Although this would certainly markedly decrease “door-to-needle” time, it also would subject patients with no LVO unnecessarily to the risks of DSA. Is the pursuit of affording more patients quicker access to endovascular intervention worth these risks? However, in this situation, revascularization inevitably would occur in patients with no salvageable penumbra. Recanalization of any LVO will likely result in more reperfusion hemorrhages. Indeed, this was reported in the series by Bhatt et al. who found a statistically greater incidence of symptomatic intracranial hemorrhage in patients with ASPECTS 0 to 5 compared to those with ASPECTS 6 to 7.12 Thus, such a triage and treatment paradigm would force practitioners to accept a poor outcome at their hands, rather than due to the natural history of a large-vessel stroke. This increased rate of hemorrhage would be “collateral damage” tolerated in the interest of expediting recanalization and maximizing the number of patients receiving mechanical thrombectomy. What imaging is necessary to triage patients presenting with stroke-like symptoms? As the authors of this article summarize, the answer to this question is variable and, at least partially, dependent on the position of the practitioner to the scenario we just described. At our institute, we were an early adopter of perfusion imaging and rely on it to guide our decision-making regarding stroke intervention. Our lengthy experience with the use of this imaging as an integral part of our triage paradigm has improved its efficiency such that we are now able to obtain noncontrast head CT, CTA from aortic arch to vertex, and CTP in less than 3 min. The radiologic software at our institute allows immediate processing of perfusion maps, such that the maps are completed and reviewed by the stroke treatment team before the patient has even left the CT scanner. By selecting patients using CTP, we believe that we are increasing the efficacy of mechanical thrombectomy (ie, more successes with fewer complications), while also upholding our oath to “first do no harm.” Although the efficiency and importance of emergency room triage of patients with acute ischemic stroke cannot be understated, expansion of the triage to “the field” recently has garnered interest. The introduction of mobile stroke units has allowed for the initiation of imaging triage,13 and in some cases treatment,14 prior to the arrival of the patient at the hospital. Prehospital triage of the patient combined with prehospital notification and coordination of treatment15 may further improve door-to-needle times and the delivery of mechanical thrombectomy. Imaging modalities that are still being explored hold promise for an innovative future. On-table imaging protocols and parametric imaging have been effective in the diagnosis of LVO,16,17 but technology to correlate this to penumbral tissue and its impact on immediate clinical decision-making is still under investigation. Because of the limited interobserver reliability in thrombectomy cases involving multiple distributions, immediate multiparametric imaging can offer a new understanding into acute intraprocedural patient assessment.16,17 Practitioners in the field continue to move forward in this direction, working to decrease recanalization times and select the best candidates for intervention. Disclosures Dr Levy is a shareholder/has ownership interests ins NeXtGen Biologics, RAPID Medical, Claret Medical, Cognition Medical, Imperative Care (formerly the Stroke Project), Rebound Therapeutics, StimMed, Three Rivers Medical; is a National Principal Investigator/on Steering Committees for Medtronic (merged with Covidien Neurovascular) SWIFT Prime and SWIFT Direct Trials; receives honoraria from Medtronic (training and lectures); is a consultant for Claret Medical, GLG Consulting, Guidepoint Global, Imperative Care, Medtronic, Rebound, StimMed; is on the advisory board for Stryker (AIS Clinical Advisory Board), NeXtGen Biologics, MEDX, Cognition Medical, Endostream Medical; and is Site Principal Investigator for CONFIDENCE study (MicroVention), and STRATIS Study—Sub I (Medtronic).

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.